Understanding the impact of pre-interview information on the reliability, validity, accuracy and differential validity of employment interview decisions: Comparisons across interview question type, rating scale and scoring protocols

Abstract

<p>The impact of pre-interview impressions on the reliability, validity, accuracy and bias of interview decisions was examined in the context of different interview types. Although it is generally well accepted that structuring the interview improves its reliability and validity, little research has been conducted with respect to resistance to interviewer bias. Dipboye (1982) proposed that pre-interview impressions can impact on post interview decisions through the interviewer's behaviour and cognitive processes. This study used videotaped interviews to assess the impact of a negative/inaccurate pre-interview impression on subsequent interview ratings via the cognitive process. Predictor data consisted of pre- and post interview ratings and criterion data consisted of supervisor ratings of 'on-the-job' performance. Three types of interview formats were manipulated: question type (job relevant versus general), rating scale (graphic versus behaviourally anchored), and combination of ratings (clinical versus mechanical). There were considerable differences in the interview results when the pre-interview information about a candidate was positive (accurate) versus negative (inaccurate). Specifically, when positive (accurate) information preceded the interview the greatest contributor to interview reliability and validity was the use of job relevant questions. However, when the interview was preceded by negative (inaccurate) information about a candidate the use of job relevant questions, behaviourally anchored rating scales and a mechanical combination of ratings were all necessary to eliminate any impact of bias on the reliability and validity of the interviewer's decision. These results indicate that structuring the interview can minimize the potentially negative effects of interviewer biases on the reliability and validity of interview decisions.</p>